[U-Boot] [PATCH] dm: mmc: socfpga: fix MMC_OPS support
Now that CONFIG_BLK and CONFIG_MMC_OPS are enabled by default with CONFIG_DM_MMC, the DWMMC driver on the socfpga platform fails at runtime.
This adds the missing fields in the driver declaration.
Signed-off-by: Sylvain Lesne lesne@alse-fr.com ---
Hi,
I might be missing something, but with the latest master, on a socfpga board, I get the following failure:
U-Boot SPL 2016.11-rc2-00072-g869588d-dirty (Oct 24 2016 - 17:18:22) drivers/ddr/altera/sequencer.c: Preparing to start memory calibration drivers/ddr/altera/sequencer.c: CALIBRATION PASSED drivers/ddr/altera/sequencer.c: Calibration complete Trying to boot from MMC1 spl: partition error MMC: block number 0xx exceeds max(0xx) mmc_load_image_raw_sector: mmc block read error SPL: failed to boot from all boot devices ### ERROR ### Please RESET the board ###
After some digging, it seems that the socfpga_dw_mmc driver slipped through the cracks during some steps of the DM conversion, so I added the missing fields of the driver declaration, mimicking what's done in the other dw_mmc_* drivers, and it works on my board.
(Also, this is kind of unrelated, but it looks like the error message 'MMC: block number 0xx exceeds max(0xx)' has a formatting issue!)
--- drivers/mmc/socfpga_dw_mmc.c | 2 ++ 1 file changed, 2 insertions(+)
diff --git a/drivers/mmc/socfpga_dw_mmc.c b/drivers/mmc/socfpga_dw_mmc.c index 5a3a4ff..0a22e58 100644 --- a/drivers/mmc/socfpga_dw_mmc.c +++ b/drivers/mmc/socfpga_dw_mmc.c @@ -151,7 +151,9 @@ U_BOOT_DRIVER(socfpga_dwmmc_drv) = { .id = UCLASS_MMC, .of_match = socfpga_dwmmc_ids, .ofdata_to_platdata = socfpga_dwmmc_ofdata_to_platdata, + .ops = &dm_dwmci_ops, .bind = socfpga_dwmmc_bind, .probe = socfpga_dwmmc_probe, .priv_auto_alloc_size = sizeof(struct dwmci_socfpga_priv_data), + .platdata_auto_alloc_size = sizeof(struct socfpga_dwmci_plat), };
On 10/24/2016 06:24 PM, Sylvain Lesne wrote:
Now that CONFIG_BLK and CONFIG_MMC_OPS are enabled by default with CONFIG_DM_MMC, the DWMMC driver on the socfpga platform fails at runtime.
This adds the missing fields in the driver declaration.
Signed-off-by: Sylvain Lesne lesne@alse-fr.com
Nice catch, thanks.
Acked-by: Marek Vasut marex@denx.de
On 24 October 2016 at 10:24, Sylvain Lesne lesne@alse-fr.com wrote:
Now that CONFIG_BLK and CONFIG_MMC_OPS are enabled by default with CONFIG_DM_MMC, the DWMMC driver on the socfpga platform fails at runtime.
This adds the missing fields in the driver declaration.
Signed-off-by: Sylvain Lesne lesne@alse-fr.com
Hi,
I might be missing something, but with the latest master, on a socfpga board, I get the following failure:
U-Boot SPL 2016.11-rc2-00072-g869588d-dirty (Oct 24 2016 - 17:18:22) drivers/ddr/altera/sequencer.c: Preparing to start memory calibration drivers/ddr/altera/sequencer.c: CALIBRATION PASSED drivers/ddr/altera/sequencer.c: Calibration complete Trying to boot from MMC1 spl: partition error MMC: block number 0xx exceeds max(0xx) mmc_load_image_raw_sector: mmc block read error SPL: failed to boot from all boot devices ### ERROR ### Please RESET the board ###
After some digging, it seems that the socfpga_dw_mmc driver slipped through the cracks during some steps of the DM conversion, so I added the missing fields of the driver declaration, mimicking what's done in the other dw_mmc_* drivers, and it works on my board.
(Also, this is kind of unrelated, but it looks like the error message 'MMC: block number 0xx exceeds max(0xx)' has a formatting issue!)
drivers/mmc/socfpga_dw_mmc.c | 2 ++ 1 file changed, 2 insertions(+)
Reviewed-by: Simon Glass sjg@chromium.org
The formatting issue may be due to tiny-printf.
- Simon
Hi,
On 10/25/2016 06:13 AM, Simon Glass wrote:
On 24 October 2016 at 10:24, Sylvain Lesne lesne@alse-fr.com wrote:
Now that CONFIG_BLK and CONFIG_MMC_OPS are enabled by default with CONFIG_DM_MMC, the DWMMC driver on the socfpga platform fails at runtime.
This adds the missing fields in the driver declaration.
Signed-off-by: Sylvain Lesne lesne@alse-fr.com
Applied on u-boot-mmc. Thanks!
Best Regards, Jaehoon Chung
Hi,
I might be missing something, but with the latest master, on a socfpga board, I get the following failure:
U-Boot SPL 2016.11-rc2-00072-g869588d-dirty (Oct 24 2016 - 17:18:22) drivers/ddr/altera/sequencer.c: Preparing to start memory calibration drivers/ddr/altera/sequencer.c: CALIBRATION PASSED drivers/ddr/altera/sequencer.c: Calibration complete Trying to boot from MMC1 spl: partition error MMC: block number 0xx exceeds max(0xx) mmc_load_image_raw_sector: mmc block read error SPL: failed to boot from all boot devices ### ERROR ### Please RESET the board ###
After some digging, it seems that the socfpga_dw_mmc driver slipped through the cracks during some steps of the DM conversion, so I added the missing fields of the driver declaration, mimicking what's done in the other dw_mmc_* drivers, and it works on my board.
(Also, this is kind of unrelated, but it looks like the error message 'MMC: block number 0xx exceeds max(0xx)' has a formatting issue!)
drivers/mmc/socfpga_dw_mmc.c | 2 ++ 1 file changed, 2 insertions(+)
Reviewed-by: Simon Glass sjg@chromium.org
The formatting issue may be due to tiny-printf.
- Simon
On 10/26/2016 04:03 AM, Jaehoon Chung wrote:
Hi,
On 10/25/2016 06:13 AM, Simon Glass wrote:
On 24 October 2016 at 10:24, Sylvain Lesne lesne@alse-fr.com wrote:
Now that CONFIG_BLK and CONFIG_MMC_OPS are enabled by default with CONFIG_DM_MMC, the DWMMC driver on the socfpga platform fails at runtime.
This adds the missing fields in the driver declaration.
Signed-off-by: Sylvain Lesne lesne@alse-fr.com
Applied on u-boot-mmc. Thanks!
Cool, thanks.
Hi,
On 10/24/2016 11:13 PM, Simon Glass wrote:
On 24 October 2016 at 10:24, Sylvain Lesne lesne@alse-fr.com wrote:
(Also, this is kind of unrelated, but it looks like the error message 'MMC: block number 0xx exceeds max(0xx)' has a formatting issue!)
drivers/mmc/socfpga_dw_mmc.c | 2 ++ 1 file changed, 2 insertions(+)
Reviewed-by: Simon Glass sjg@chromium.org
The formatting issue may be due to tiny-printf.
- Simon
Thanks for the review! And indeed, I didn't realize that tiny-printf doesn't support the 'long int' specifiers, sorry.
-- Best regards, Sylvain Lesne
participants (4)
-
Jaehoon Chung -
Marek Vasut -
Simon Glass -
Sylvain Lesne